Further studies are needed to understand the root causes low recruitment levels, including pollinators and pollination success studies, climate change impact studies, seed and seed bank viability studies, and genetic studies to analyze for potential deleterious effects caused by inbreeding depression.

Low levels of recruitment combined with herbivory on plants and seeds appear to be the largest threats to the natural and experimental populations.

Whether establishing new populations of Pediocactus knowltonii by transplanting clones or direct seeding can be successfully accomplished has yet to be seen. Natural recruitment to these new populations has been an exceedingly slow process. Overall, it appears that direct seeding large quantities of seeds is more likely to succeed in establishing self-sustaining populations than those started from clones. A population viability analysis might shed some light on what constitutes a viable population and how many plants are needed for successfully establishing a new population.

In addition, it would be helpful to know what percentage of Pediocactus knowltonii seeds are viable and therefore contribute to the viability of the seed bank. If a large percentage of seeds are not viable, what is the cause? Lack of pollinators or pollination success? Inbreeding depression? Similarly, whatpercentage of seeds are carried off by seed predators and therefore become lost to the population?

the percentage of plants in the juvenile size class remains significantly higher in the seed plots over the natural population and transplanted populations. Although this could be an artifact of detection, it is possible that at least some of the original 1994 seeds may continue to contribute to the population of newly recruited individuals in the BLM seed plots.

Direct seeding into new locations is a viable alternative to transplanting adult clones. However, only about 5% of the seed becomes established as adult plants and they require a longer period to become reproductive than do transplanted clones. A large quantity of seed is required to ensure adequate germination and establishment. Recruitment from the seedbank remains highest in the seed plots, but in general, seed plot plants are moving towards a natural size class distribution similar to the natural population. After 7 years, the small population in the seed plots at the BLM #1 location was stable with recruitment roughly equal to mortality until 2007 and 2008 which had net declines in number of cacti. Slight increases in numbers from 2009 through 2012 appeared to constitute a small, but stable or growing, population. However, plant numbers dropped sharply in 2015. The subsequent increase in plant numbers and high number of seedlings may be the result of increased spring rainfall in 2015.

Although some new plants have been recorded since 1986, seedlings have rarely been documented and the transplant site continues to retain the lowest percentage of plants in the juvenile age class.

uffered a catastrophic decline caused by rodent predation in 2006 and the transplant population was abandoned and judged a failed effort in 2007.

27 CONCLUSIONSAlthough the number of Pediocactus knowltonii plants in the monitoring plots fluctuates from year to year, the overall trend after 31 years of monitoring at the type locality is a slow decline and was especially pronounced in 2015. This trend is consistent with other species of Pediocactus monitored on the Colorado Plateau (Clark & Clark 2008, Hazelton 2011, Phillips & Phillips 2004, Roth 2008, USFWS 2012). Declining trends are largely attributed to prolonged drought impacts associated with global climate change. Climate change impacts may include changes in pollinator availability and therefore successful pollination among Pediocactus plants, increase in predation by beetles, rabbits, and rodents, desiccation, decreased reproductive effort and therefore decreased germination and establishment. In addition, climate change is expected to shift ecological functions and timing, including shifts in flowering time, vegetation community changes, and changes in pollinator and herbivore community composition. Increased mortalities may occur during years of higher than usual rainfall due to increased resource competition withseedlings of annual invasive species and increased predation by herbivores due to increases in their population numbers with the availability of additional forage. Although no invasive species were documented at the Sabo Preserve, increased predation is thought to be the primary cause of the decline 2015. The natural population increased in numbers during the late 1980s and early 1990s, and then gradually declined to the lowest point recorded in 2017. It was estimated that the peak population in 1994 would be about 14,000 cacti, if the 1992 estimate of 12,000 plants was accurate (USFWS 2010). By this same 1992 benchmark, the monitoring plot data suggested a total population of only 6,100 cacti in 2008 (Sivinski 2008). A thorough inventory of plants in 2015 found only about half that number at the Sabo Preserve. Additional surveys in suitable habitat to potentially document new natural populations were unsuccessful, supporting our current understanding that the total world population of Pediocactus knowltonii is restricted to one small hill, providing habitat for less than 3,500 plants. This decline is likely a combination of low reproduction and recruitment, influenced by ongoing long-term drought conditions, seed and plant predation from rodents, rabbits, and potentially insects. Other factors contributing to the decline might include small population size and associated inbreeding depression, pollinator availability and pollination success.One serious episode of cactus poaching was detected in 1996 when an entire monitoring plot and an undetermined number of cacti were removed from the natural population at the type locality. No further acts of vandalism have been identified since, although it continues to be a viable threat to Pediocactus knowltonii due to its collection value and overall rarity. After a steady decline of transplants over the initial 10-year period, followed by the loss of 2 transplant sites due to rodent predation, the BLM #1 transplant site appears stable. The multi-stem donor plants in the natural population did not suffer from the loss of a single stem. However, transplanted clones comprise an aging population with little recruitment. These plants are relatively long-lived for small cacti,

27 CONCLUSIONSAlthough the number of Pediocactus knowltonii plants in the monitoring plots fluctuates from year to year, the overall trend after 31 years of monitoring at the type locality is a slow decline and was especially pronounced in 2015. This trend is consistent with other species of Pediocactus monitored on the Colorado Plateau (Clark & Clark 2008, Hazelton 2011, Phillips & Phillips 2004, Roth 2008, USFWS 2012). Declining trends are largely attributed to prolonged drought impacts associated with global climate change. Climate change impacts may include changes in pollinator availability and therefore successful pollination among Pediocactus plants, increase in predation by beetles, rabbits, and rodents, desiccation, decreased reproductive effort and therefore decreased germination and establishment. In addition, climate change is expected to shift ecological functions and timing, including shifts in flowering time, vegetation community changes, and changes in pollinator and herbivore community composition. Increased mortalities may occur during years of higher than usual rainfall due to increased resource competition withseedlings of annual invasive species and increased predation by herbivores due to increases in their population numbers with the availability of additional forage. Although no invasive species were documented at the Sabo Preserve, increased predation is thought to be the primary cause of the decline 2015. The natural population increased in numbers during the late 1980s and early 1990s, and then gradually declined to the lowest point recorded in 2017. It was estimated that the peak population in 1994 would be about 14,000 cacti, if the 1992 estimate of 12,000 plants was accurate (USFWS 2010). By this same 1992 benchmark, the monitoring plot data suggested a total population of only 6,100 cacti in 2008 (Sivinski 2008). A thorough inventory of plants in 2015 found only about half that number at the Sabo Preserve. Additional surveys in suitable habitat to potentially document new natural populations were unsuccessful, supporting our current understanding that the total world population of Pediocactus knowltonii is restricted to one small hill, providing habitat for less than 3,500 plants. This decline is likely a combination of low reproduction and recruitment, influenced by ongoing long-term drought conditions, seed and plant predation from rodents, rabbits, and potentially insects. Other factors contributing to the decline might include small population size and associated inbreeding depression, pollinator availability and pollination success.One serious episode of cactus poaching was detected in 1996 when an entire monitoring plot and an undetermined number of cacti were removed from the natural population at the type locality. No further acts of vandalism have been identified since, although it continues to be a viable threat to Pediocactus knowltonii due to its collection value and overall rarity.

This decline is likely a combination of low reproduction and recruitment, influenced by ongoing long-term drought conditions, seed and plant predation from rodents, rabbits, and potentially insects. Other factors contributing to the decline might include small population size and associated inbreeding depression, pollinator availability and pollination success

Additional surveys in suitable habitat to potentially document new natural populations were unsuccessful, supporting our current understanding that the total world population of Pediocactus knowltonii is restricted to one small hill, providing habitat for less than 3,500 plants

radually declined to the lowest point recorded in 2017. It was estimated that the peak population in 1994 would be about 14,000 cacti,

increased predation is thought to be the primary cause of the decline 2015.

the overall trend after 31 years of monitoring at the type locality is a slow decline and was especially pronounced in 2015. This trend is consistent with other species of Pediocactus monitored on the Colorado Plateau (Clark & Clark 2008, Hazelton 2011, Phillips & Phillips 2004, Roth 2008, USFWS 2012). Declining trends are largely attributed to prolonged drought impacts associated with global climate change. Climate change impacts may include changes in pollinator availability and therefore successful pollination among Pediocactus plants, increase in predation by beetles, rabbits, and rodents, desiccation, decreased reproductive effort and therefore decreased germination and establishment. In addition, climate change is expected to shift ecological functions and timing, including shifts in flowering time, vegetation community changes, and changes in pollinator and herbivore community composition.

Mortality has consistently outpaced recruitment over the entire 31 years of monitoring.

long-term drought impacts, seed and plant predation, some illegal collecting, and potential impacts from inbreeding depression resulting in low fecundity and therefore declining seed deposits into the seedbank.

The current estimate of the total number of plants in existence documents an even more precipitous decline (70%).

This represent a significant decline (73%) over the 1992 estimate of 12,000 plants.

he 2015 full inventory of the natural population counted a total of 3,445 individuals at the TNC Sabo Preserve.

In 2005, eleven years following seeding and plot treatment, an analysis of variance for the random block design of this experiment showed no significant differences in the number of plants between plot treatments

The percentage of plants reproducing in the BLM seed plots increased steadily since and even surpassed the percentage of reproducing plants in the natural population between 2005 and 2011. I

he percentage multi-stemmed individuals remains lowest in the seeded plots among the three monitored populations (21% in 2017) (Figure 8).

wltonii plants started from 2,250 seeds at 3 BLM seed plots in San Juan County, NM.

In 2017, 99 plants were found within all three seed plots (Figure 9). This represents a 14% increase in numbers over the 2016 resultsand is likely associated with germination and establishment in response to increased spring rainfall in 2015 and represents a recovery from the declines observed in 2015. The majority of plants were rated in excellent or good condition (Figure 10)

This was likely related to the drought of 2001/2002 in combination with an exhaustion of the artificially stocked seedbank (Figure 1)

Only 4 of the original 12 survived to be counted again in May 1995. A total of 69 new germinants were counted in the 1995 assessment. The seedlings were not readily visible during the severe drought year of 1996 and a complete assessment was not made during that year.

The BLM #1 transplant site was also seriously impacted by predation in 2006 and again in 2007

he late season planting also did not allow sufficient

Unlike the Navajo #1 & #2 sites, these cacti were planted bare-root and may have lacked the additional anchor of artificial potting soil.

The Navajo #1 and #2 transplants slowly dwindled away, without significant recruitment. The entire transplant population catastrophically declined in the winter of 2005/2006, when rodent or rabbit predation killed most of the plants remaining at the Navajo Lake transplant location (Figure 6) (Sivinski 2006).

resulting in a 12 percent rate of mortality from May 1991 to May 1992. During this same period of time, unmolested, multiple-stemmed cacti in adjacent study plots experienced a natural mortality rate of 12.9 percent (N= 101). Therefore, no increase in mortality resulted from the stem damage incurred during the cloning operation

Since the original planting in 1991, new plants have established along the transect lines in various locations

The Navajo Lake plants were transplanted with the rooting medium still attached to the roots. The BLM #1 transplants were entirely bare-root plantings.

In September 1991, a total of 149 five-month-old clones where planted on the BLM's Reese Canyon Area of Critical Environmental Concern (ACEC), which is referred to as the BLM #1 Site

An additional 250 cuttings were taken in the spring of 1991 (Sivinski and Lightfoot 1992)

The reintroduction (transplant) program began in May 1985, when 250 stem cuttings were taken from multi-stem plants at the type locality (Olwell et al. 1987). These clones were taken to a greenhouse and grown in pots over the summer until fully rooted. One hundred fifty of these adult clones were planted at the transplant location adjacent to the Los Pinos arm of Navajo Lake, which is hereafter referred to as the Navajo #1 Site. They were planted in fall of 1985 in a grid pattern at two-meter intervals along 15 lines of ten plants each. This site was supplemented with another 102 cuttings planted on the south side of this grid in the early spring of 1995 (Sivinski 1995). These later transplants are in the same general area, but are referred to as the Navajo #2 Site

In 2017, 30% of live plants in the monitoring plots were multi stemmed

may begin to develop multiple stems in response to disturbance

The majority of individuals are young, reproducing adults, with a diameter of 1.1 to 2.0 cm ( 36% in 2017) (Figure 8). In 2017 approximately 11% of all plants in the natural population were juvenile, non-reproducing plants less than 1.0 cm in diameter (excluding seedlings).

In 2017, the mean diameter of Pediocactus knowltonii in the natural population was 1.84 cm

but was lowest during the drought years of 1996 and 2002 (Table 2).

The highest number of seedlings recorded was in 2017 (16 seedlings). It is likely that the germination and establishment of these seedlings are the result of the high rainfalls in 2015, or possibly the results of higher than average rainfall in the winter of 2016/2017. However, recruitment continues to lag significantly behind mortality in the natural population throughout the monitoring years

Prolonged drought conditions and the associated lower levels of reproduction and recruitment are generally considered the likely driving force behind the decline of this population. In addition, dry conditions can cause an increase in rabbit and rodent attacks, which are frequently fatal.

The highest number of plants reported missing, gone, or dead since 2012 were recorded in 2015 (137 plants) and 2016 (110 plants).

it is estimated that individual Pediocactus knowltonii plants can live for at least 30 years in their natural habita

Reproductive effort was lowest during the drought years of 1996 and 2002 (Figures 1 & 5)

The majority of plants were in excellent or good condition (Figure 3)

In 2015, only 138 plants were found within the ten monitoring plots. This represents the sharpest decline in the number of plants documented since 1984 and a 37% decline from 2014 when 219 plants were found in the 10 monitoring plots.

7 might reflect the establishment of seedlings during the unusually high rainfalls in the winter of 1986/198

A similar drought during 2001/2002 is considered the main cause of a steep decline in plant numbers in 2002 (Figures 2 & 4). The number of plants found in the monitoring plots decreased by 25% from 2001 to 2002

Therefore, the trend between 1995 and 1997 might be more gradual than shown in Figure 4.

The dry winter of 1995 to 1996 corresponded to a steep decline in plant numbers, which dropped by 27% within one year(Figures 2 & 4).

ave dropped significantly in 2015 (37% over 2014 numbers), with only a slight increase documented in 2016, followed by an additional decline in 2017 (11% over 2016 numbers).

Although the population trend initially increased by 78% between 1986 and1994, it has decreased since 1995 to a number significantly below the original 1986 density

verall, the number of juvenile and adult plants within the 10 monitoring plots at the type locality has been declining over the past 31 years and reached an all-time low in 2017 (Figure 4).

)

Annual monitoring is conducted during the first 2 weeks of May at Type Locality/Natural Population:1.the TNC Sabo Preserve (1986 to present)Transplant Sites:2.the BLM #1 Transplant Site (1991 – present), 3.the BLM #1 Seed Plots Site (1994 – present)

The average annual precipitation at Aztec National Monument ( ca. 25 miles SW of the type locality) is approximately 11 inches, ranging from approximately 3 to 20 inches over a 97-yearperiod (WRCC 2018) (Figure 1).

Approximately one-third of the natural population has 2-15 stems per plan

This small cactus has contractile roots, which can pull the entire plant below the soil surface during periods of severe drought

A full inventory of the Sabo Preserve documented approximately 3,500 plants total in 2015 (Roth 2015).

This cactus grows at about 6,200ft, in full sun or partial shade, between ancient river cobble

ediocactus knowltonii habitat occurs on Tertiary alluvial deposits overlying the San Jose Formation.

high as 100,000 plants in 1960, then was reduced by cactus collectors to only 1,000 plants by 1979

a peak in 1994 at 17% higher than 1992. Therefore, the natural population was estimated at about 14,000 cacti in 1994 and has been gradually and steadily declining since then.

. A reintroduction program into nearby suitable habitats was identified as the primary effort towards recovery of this species.

In an effort to protect the only natural population of this rare cactus, the original landowner (Public Service Company of New Mexico) donated the 10-hectare type locality to The Nature Conservancy (TNC).

his population was severely impacted by the New Mexico Cactus and Succulent Society in 1960, which was under the mistaken perception that this site would be flooded by the newly constructed Navajo Reservoir (USFWS 1985). Field trips were organized to salvage the cacti from the type locality. Several thousand Pediocactusknowltonii plants were reportedly taken by this group of hobbyists

Pediocactus knowltonii L. Benson (Knowlton's cactus) is one of the rarest cacti in the United States. It was discovered in 1958 by the late Fred Knowlton and named by Lyman Benson in 1961. It was listed endangered by U.S. Fish & Wildlife Service (USFWS) on October 29, 1979 (44 FR 62244).

It was discovered in 1958 by the late Fred Knowlton and named by Lyman Benson in 1961

To help design an optimal management plan, we used demographic data from a 5-year fieldstudy of H. montana—which included a controlled burn—to construct size-based population projection matri-ces. Using these matrices, we projected the consequences of instituting controlled burns and hiking and camp-ing restrictions separately and in tandem. We also determined the burn frequency that would maximize H.montana’s population growth rate. Finally, we used a stochastic model to determine how environmental fluc-tuations could alter the efficacy of conservation measures.

A two-pronged manage-ment strategy is more apt to succeed in part becauseburning makes cliff-side habitat more accessible to hikersand campers.

ur results suggest that management strategies that in-tegrate controlled burning with trampling reduction holdmore promise. Combining the two management tacticscould yield a deterministic growth rate of 1.0152 and an88.75% chance that the H. montana population willgrow over the next 200 years

oth controlled burns and trampling reduction can slowH. montana’s decline, but neither strategy by itself is suf-ficient to produce a growing H. montana population.

When combined, these two management tacticsyield higher population growth rates than either manage-ment tactic alone.

The stochastic model also predicted thatH. montana populations would increase without burn-ing if trampling were eliminated, whereas the determinis-tic model did not. The higher population growth rates inour stochastic model support speculation (Frost 1990)that the occasional benefits of wet years are essential tomaintaining viable H. montana populations.

Contrary to the deterministic model, thestochastic model predicted that H. montana populationgrowth was possible with only a 50% reduction in tram-pling mortality.

Controlled burning at short and intermediateintervals caused an increase in the proportion of individ-uals in the three larger size classes, whereas burning atlong intervals produced only a minor shift in the stablesize distribution

Both models predicted that combining high lev-els of trampling reduction with burn cycles of intermedi-ate length would offer the most effective managementstrategy.

1298Conservation Models for H. montanaGross et al.Conservation BiologyVolume 12, No. 6, December 1998(z10%) higher sensitivities to changes in uncertain param-eters than did l for the control matrix. In no case did a 1%perturbation of an uncertain parameter change the opti-mal burn frequency for a given level of trampling control.Stochastic ModelThe stochastic model predicted effects of burning andtrampling reduction similar to those predicted by the de-terministic model. Greater trampling reduction increasedthe average population growth rate (Fig. 2b) and the frac-tion of simulations with growing populations (Fig. 5).With reductions in trampling mortality, population growthrates were maximized at intermediate burn frequenciesthat roughly coincided with the optimal burn frequenciespredicted by the deterministic model

Annual population growth rate as a func-tion of burn cycle length and trampling reduction for deterministic (a) and stochastic (b) models. In (a) solid lines are asymptotic rates and dashed lines show average annual populatio

Conservation BiologyVolume 12, No. 6, December 1998Gross et al.Conservation Models for H. montana1297was eliminated entirely.

the benefits of burning are maximized whenthe interim between burns is long enough to permit fullH. montana recovery yet not long enough to allow L.buxifolium to recover as well.

For all three levelsof trampling reduction, population growth rates peakedat burn cycle durations of 5–8 years and asymptoticallyapproached the control growth rate as burn frequencydecreased

Although both controlled burns and reduction in tramplingmortality increased H. montana’s population growth rate,neither tactic was sufficient by itself to produce a grow-ing H. montana population

Specifically,we projected H. montana’s population growth rate un-der management plans incorporating either controlledburns, trampling reduction, or both. We also modeled theresponse of H. montana populations to a range of burnfrequencies. Because environmental fluctuations signifi-cantly affect the shrub’s demographic rates, we con-structed two complementary models: a deterministicmodel for tractable analysis of the effects of managementtactics and a stochastic model to explore the impact of acapricious environment on management success.

First, what might be the relative contributions ofcontrolled burns and trampling reduction in aiding re-covery of H. montana? Are both tactics necessary, orcould one or the other suffice?

There are norelevant conservation units in this region. The highgenetic diversity and uniqueness found in these popula-tions ofD. nigra, plus evidence from animal studies(Ditchfield, 2000; Costa, 2003; Moraes-Barroset al., 2006;Tchaickaet al., 2007), indicate the high priority thatshould be given to implementing conservation units innortheast Minas Gerais

These genetic data, high-lighting the uniqueness of populations in each phylogeo-graphic group must also be considered upon samplingseeds for theex situconservation ofD. nigra

The populations RDO and VRD, located in largeconservation units in the SG1 group, are among thosethat most contributed to the total diversity ofD. nigra.This fact shows the importance of preserving large areasto maintain genetic diversity.

we suggest that the three phylogeographic groupsofD. nigracould be treated as different MUs

These molecular data areevidence that timber exploration and deforestation havepromoted genetic depletion in this species, and also thatthe present diversity may represent only a part of thegenetic diversity seen in the past in this wide geogra-phical area of the Atlantic Forest. This scenario found forD. nigracould probably be extended to other timberspecies of the Atlantic Forest.

allelic richness corrected for the sample size

In addition to the signature of the effects of pastclimatic changes

can be suggestedthat the northernmost populations (NG and SG1)maintained large populations during a great periodand that the SG2 may have undergone a morepronounced bottleneck.

Furthermore, the NG andSG1 showed higher effective population sizes than theSG2. Because the effective size is a harmonic average ofthe effective size across generation

hat is, after the splitthere were population expansions. The NG and SG1exhibited high diversities (Table 3), whereas the south-ernmost group (SG2) had low diversity, as expected for arecently founded population.

ur analysis showed evidence of weak populationexpansion in the NG and SG1

These datashowed that the drier conditions of the LGM causedthe Atlantic Forest to become much smaller and morefragmented than it is today (Behling, 2002). As a result ofthis constriction, the rain forest was probably partlyreplaced by semi-deciduous forest in southeastern Braziland by semi-deciduous forest and caatinga (tropicalsemi-desert shrub vegetation) in northeast Brazil (Behl-ing, 1998, 2002). A palynological study in the region ofthe split region between the Doce and Jequitinhonharivers, performed by Behling (1995) in Lago do Pires(171590S, 421130W) showed that this region in the earlyHolocene was dominated by species from Cerrado (theBrazilian savannah), suggesting the occurrence of longdry seasons. This study also suggested that the currentdense and closed semi-decidous forest in this region wasonly established in the latest Holocene (beginning 970years ago). The recurrence of climatic changes during thePleistocene in the central part of the Atlantic forest, withcycles of forest expansion and contraction, likely led torepeated vicariance events and thus resulted in thegenetic differentiation of the three population groups ofD. nigra. In fact, the location of the current NG group isincluded in the putative Bahia refugium, a large stableregion from the Doce River northward to the southernborder of the Sa ̃o Francisco River described by Carnava

Considering only the genetic divergences among thegroups, the estimated divergence times between the NGand SG1 groups and between the SG1 and SG2 groupswere similar (351 000 and 389 000 ybp, respectively). Weapplied the IM analysis taking into account demographicevents because of possible differences in the effectivepopulation sizes between populations and gene flowafter split. This analysis showed a more ancient separa-tion time between these groups, 612 000 ybp (NG/SG1groups) and 784 000 ybp (SG1/SG2). In spite of thedifferences between the estimates with the two ap-proaches, both are concordant in pointing to vicarianceevents in the Middle Pleistocene.

D. nigra, between the SG1 and SG2 groups, was locatedaround 211S, in the same region of separation betweengenetic groups found in another two legume trees,Caesalpinia echinata(Liraet al., 2003) andHymenaeacourbaril(Ramoset al., 2009). However, in thePodocarpusspecies (Ledruet al., 2007), the separation among thethree groups appears not to coincide with the split zonesfound in legume trees. These latitudinal genetic disjunc-tions found in different groups of species point tohistorical vicariance events in the central Atlantic Forest.

The genetic splitseparating the NG and SG1 groups ofD. nigrawas foundaround 18oS latitude, between the Jequitinhonha andDoce rivers, coinciding with the split zone range foundin several animal species

identified as threephylogeographical groups, the northern (NG), southern1 (SG1) and southern 2 (SG2) groups (Figure 2)

62% of the total genetic diversity due to differencesamong populations (FST¼0.624)

he divergence time betweenthe main groups, NG and SG1, was estimated at612 595 ybp (lower bound of HPD 90% interval:372 137). The two groups of southern region, SG1 andSG2, diverged earlier, at 784 351 ybp (lowe

NG and SG1diverged approximately 351 000 years before the present(ybp) (CI¼323 092–1 025 072). The time of divergencebetween SG1 and SG2 (dA¼0.00079; T¼389 000 ybp;CI¼400 538–1 178 953) was similar to that between NGand SG1

abbreviations and molecular diversity indexes. The median-joining network analysis is represented on the upper left. The size of each circle

TheNST(0.649±0.0624;Po0.05)was significantly higher than theGST(0.559±0.0690),indicating a phylogeographic structure, that is, closelyrelated haplotypes in theD. nigrapopulations werefound more often in the same area than less closelyrelated haplotypes.

he multiple regression analysis indicated that ap-proximately 69% of the variation (R-squared) in thegenetic distances could be predicted by the geographicallocation (partial correlation coefficient¼0.463,Po0.01)and the linear geographical distance among localitypairs (partial correlation coefficient¼0.689,Po0.01).This result indicates that both vicariance processes andisolation by distance explain the present-day geneticstructure ofD. nigra.

the Monmonier algorithm (BARRIERpro-gram) suggested the following main barriers to gene flowin theD. nigradistribution range: (1) one barrier locatedapproximately in the northeast of Minas Gerais andsouth of Bahia, dividing the natural distribution of thisspecies into northern and southern groups, and (2)another barrier located between Espirito Santo and Riode Janeiro, thus isolating theD. nigrapopulations locatedin the extreme south of the sampling area (Figure 2).

The AMOVA analysis showed a very strong differ-entiation among allD. nigrapopulations (FST¼0.624,Po0.0001). In theSAMOVAanalysis, the best groupingscheme divided the distribution range ofD. nigrai

The comparisonof the genetic diversities of populations from smallfragments (TOT, SCR, DIO, ACA, PMI and VNI) withthose from large reserve areas (RDO and VRD; Figure 2),both pertaining to same geographical group (see belowthe results ofSAMOVAprogram), showed significantlylower genetic diversities in all estimates in the smallfragments, such ash(t¼6.57,Po0.001),p(t¼8.57,Po0.001) andHr(t¼8.13,Po0.001).

15 haplotypes with 11polymorphic sites,

(1) What is the influence of the historical events of thePleistocene on the current genetic diversity and structureofD. nigra, and specifically, are there genetic disjunctionsalong its distribution? (2) In addition to historical events,were the genetic diversities of populations affected byhuman-mediated fragmentation?

studies that take into account bothof the factors that affect the current genetic structure ofspecies—historical and human-mediated events—arerequired to better understand the evolutionary processand establish effective conservation measures

Additional phylogeographic studieswould help to reconstruct the evolutionary history ofthe central Atlantic Forest, improving conservation andmanagement measures such as the identification ofpriority populations/areas for reserve implementation.

Although thestudies in plants have attributed these genetic splits toclimatic changes of the Quaternary, none has estimatedthe time since splitting.

Based on comparisons among thepopulations of large reserves and small, disturbed fragments ofthe same phylogeographic group, we also found evidence ofrecent anthropogenic effects on genetic diversity. The resultswere also analysed with the aim of contributing to theconservation ofD. nigra. We suggest that the three phylogeo-graphic groups could be considered as three distinct manage-ment units. Based on the genetic diversity and uniqueness ofthe populations, we also indicate priority areas for conservation.Heredity(2011)106,46–57;doi:10.1038/hdy.2010.64;published online 2 June 2010

he population diversity,phylogeographic structure and demographic history of thisspecies were investigated

making them more susceptible to stand-­‐replacing fires which can lead to novel ecosystems on new successional trajectories away from historical stand characteristics

ver a century of fire suppression in warm/dry mixed conifer forests has shifted species composition toward more mesic, shade tolerant species such as white fir and Douglas-­‐fir, increased tree density, and increased surface and aerial fuels

Historically, fires in cool/moist mixed conifer forests burned at sub-­‐decadal to century frequency with a mixed-­‐severity fire regime, where surface and crown fire behavior could occur during the same fire event (Margolis and Balmat, 2009; Romme et al., 2009).

Surface fires were frequent before the mid to late 1800s depending on the site, burning with multi-­‐to sub-­‐decadal frequencyin the warm/dry mixed conifer

The root samples were placed into a 2-ml screw-cappropylene tube together with two tungsten carbide balls (3 mm) andground (3 min, 13,000 rpm) using a mixer mill (MM 400; Retsch, Haan,Germany).

when the conditions are favorable (rainy periods duringwinter and spring)

The application of molecular techniques in recent years toidentity AMF in the field, particularly in plant roots, has revealeda high diversity, including many sequences that cannot be relatedto known taxa (19,30)

Previous studies have revealed that AMFstimulate the growth of shrubs and improve their drought toler-ance (15,33), while at the same time, the shrubs exert a selectivepressure on the AMF species

Plant diversity and productivity in ecosystems are influencedsignificantly by the AM fungal diversity in the soil

Rarefaction curves showed that the number of sequences inthe 12 samples obtained from coconut roots was sufficient toasymptote, indicating that an acceptable mycorrhizal richnesswas recovered

Mycorrhizae also promote plant growth and leadto differential effects in growth rates in coconut seedlings, aswell as suppressing plague infections, depending on the AMFspecies

nd are considered among the most ecologically im-portant soil microorganisms due to the role they play inecosystem functionality, increasing plant biomass and im-proving resistance to moderate salinity and hydric stress aswell as tolerance to some pathogens

Arbuscular mycorrhizal fungi (AMF) belong to the sub-phylumGlomeromycotina, in the phylumMucoromycota(Spatafora et al.2016). Species of this group are associatedwith the roots of over 80% of land p

rhizosphere soil

The diversity and community structure of arbuscular mycorrhizal fungi (AMF) associated with coconut (Cocos nucifera)rootswas evaluated by next generation sequencing (NGS) using partial sequences of the 18S rDNA gene

No indicator species of known EM fungi wereidentified for the burned sites in either range or for the unburnedareas in the Santa Catalina Mts.

Species accumulation curves (Fig. S2) indicate that oursampling captured the majority of EM fungal species in sites rep-resenting each range and the recent history offire

EM fungi known from indi-vidual ranges may be lost as these insular forests experience morefrequent and more intense wildfires

These forests, sub-ject to drought, warming temperatures, andfire, have the potentialto complement studies conducted in California and the PacificNorthwest for a broader understanding of EM responses tofire inPonderosa pine forests

However, in isolated forests or for-est fragments, the capacity to recolonize from less- or undamagedforests may be restricted.

However, more intense and frequentfires can destroy the viable spore bank, hindering re-establishmentof fungal communities and subsequent recolonization by pines andother tree species

Afterfire, remnant communities of EM fungioften are dominated by species from the spore bank, which usuallyare disturbance-adapted and poor competitors

Such organisms include soil-inhabiting fungi,which play an integral role in forest health, productivity, andresilience to environmental stress

availability of coop-erative plant hosts and stabilization of the soil chemistryprompts the recovery of mycorrhizal community com-position and phylogenetic structure

Prior work has shown how AMF com-munities are surprisingly resilient to disturbance and canrecover quickly in the absence of dispersal, which sug-gests the presence of a local‘microbial seed bank’

AMF community resilience depends on thepost-fire regrowth of understory vegetation and the sub-sequent recovery of soil chemical properties

Shannon index andevenness were positively correlated with herbaceousbiomass and shrubby biomass, and negatively correlatedwith AP and NH4+(P<0.05 in all cases; Fig.4;TableS6). Similarly, VT richness was positively corre-lated with herbaceous biomass and shrubby biomass,and negatively correlated with AP, NH4+and pH

AMF communities were phylogeneti-cally clustered

Fire greatly reduced the abundance ofDiversisporaceaeand increased the abundance ofAcaulosporaceae, while the relative abundances ofthese taxa returned to a similar level as found in thecontrols after 11 years

). The ses.MNTD can be used to test forphylogenetic clustering or overdispersion

Mean nearest taxon distance(MNTD) (Webb2000) was calculated to evaluate theeffect of fire on the AMF phylogenetic structure usingthe picante package

The relationships between the alpha-diversity and geo-chemical features were assessed with linear regressionanalyses using SPSS 20 for Windows.

), poor-quality (below an av-erage quality score of 25) and short (<200 bp) sequenceswere removed and sequences were demultiplexed.Sequences were clustered into OTUs (operational taxo-nomic units) using a 97 % identity threshold (defaultQIIME settings) by UCLUST (Edgar2010 ). The mostabundant sequence within each cluster was selected asthe representative sequence for that OTU. To assign theOTUs to AMF virtual taxa (VT), sequences were blast-ed against the MaarjAM Glomeromycota database(Öpik et al.2010 ), which correspond to a similarity≥97 %. We rarified the abundance matrix to 280 se-quences per sample to obtain relative abundances. Thenon-AMF sequences were assigned to OTUs (97 %identity)

uantifiedusing NanoDrop ND-1000

Understory vegetation was assessed in five 2×2 mareas, one in the center and four at the corners of the40×40 m plot.

All the samples were sieved, thor-oughly mixed and divided into two parts: one part wasstored at 4 °C for biogeochemical analysis; the other wasstored at−40 °C for DNA analysis.

how do wildfires of varying intensityalter soil AMF community composition, diversity, andphylogenetic structure

After a fire, AMF communities likely help me-diate the physical stabilization of soils, the recovery ofdamaged plants, and the recruitment of new vegetation

). Microbescan be killed directly by heat, or indirectly by shifts insoil chemistry and vegetation. In addition, wildfire dis-turbance affects soil microbial phylogenetic structure(Ferrenberg et al.2013). For example, some studieshave reported that disturbance promotes phylogeneticoverdispersion (Didham et al.2005; Didham andNorton2006 ), which may indicate increased competi-tion among related taxa with overlapping niches

. Previous studies haveshown how wildfire reduces belowground microbialbiomass (Stendell et al.1999;Vernesetal.2004;Wang et al.2012), mycorrhizal fungal root-colonization

Fires oftenmineralize a significant fraction of soil organic matter(Choromanska and Deluca2001 ; Fernández et al.1997 )and reduce plant nutrient uptake (Smithwick et al.2009), resulting in a pulse of ammonium (inorganicNH4+) and available phosphorus (AP) (Neary et al.1999 ; Certini2005 ; Wang et al.2012 ). The reactivecharcoal layer left behind by wildfire can influencemany other soil properties, like pH and C availability(Wardle et al.1998 ;DeLucaandSala2006 ). Wildfiresalter the vegetation structure by selectively killingsmall-diameter trees and enhancing the growth ofgrasses

using a wide range of methods, including comparative sequence analysis, site-directed mutagenesis, genetics, mRNA profiling, co-immunoprecipitation, and proteomics, have repeatedly shown that perfect pairing to miRNA nucleotides 2–7, known as the miRNA seed, is important for the recognition of many if not most miRNA targets3. To impart more than marginal repression of mammalian targets, this seed pairing is usually augmented by either a match to miRNA nucleotide 8 (7mer-m8 site)4–7 or an A across from nucleotide 1 (7mer-A1 site)4,7 or by both (8mer site)4,7. In relatively rare instances, targeting also occurs through 3′-compensatory sites4,5,8 and centered sites9, for which substantial pairing outside the seed region compensates for imperfect seed pairing.A single miRNA can target hundreds of distinct mRNAs through seed-matched sites10. Indeed, most human mRNAs are conserved regulatory targets8, and many additional regulatory interactions occur through nonconserved sites11–13. However, not every site is effective; 8-nucleotide sites are more often effective than 7-nucleotide sites, which are more often effective than 6-nucleotide sites7,14. Another factor is site context. For example, sites in the 3′UTRs are more often effective than those in the path of the ribosome7. Among 3′UTR sites, those away from the centers of long UTRs and those within high local A–U sequence context are more often effective7, consistent with reports that sites predicted to be within more accessible secondary structure tend to be more effective15–19. Also influencing site efficacy is proximity to other miRNA-binding sites7,20, to protein-binding sites21, and to sequences that can pair to the 3′ region of the miRNA, particularly nucleotides 13–17 (ref 7).Studies of site efficacy have focused primarily on different sites to the same miRNA, without systematic investigation of whether some miRNA sequences might be intrinsically more proficient at targeting than others. Broadly conserved miRNAs typically have many more conserved targeting interactions than do other miRNAs4,8, and highly or broadly expressed miRNAs appear to target more mRNAs than do others22, but these phenomena reflect evolutionary happenstance more than intrinsic targeting proficiency.Our interest in targeting proficiency was spurred by intriguing results regarding the lsy-6 miRNA. When tested in C. elegans, only one of 14 predicted targets with 7–8-nucleotide seed-matched sites responds to lsy-6, which was interpreted to show that perfect seed-pairing is not a generally reliable predictor for miRNA–target interactions23. An alternative interpretation, which seemed more parsimonious with findings for many other miRNAs in other contexts3, is that the results for lsy-6 might not be generally applicable to other miRNAs because lsy-6 might have unusually high targeting specificity because of unusually low targeting proficiency.

n some cases, there hasbeen evidence that post-fire forest management techniques can createadditional site disturbance and enhance post-fire runoff, erosion, andsediment delivery to streams

Common post-fire forest manage-ment approaches include salvage logging, emergency stabilization, sub-soiling (cutting furrows along the contour of hillslopes), contour-felledlogs, application of straw wattle to hillslopes, and seeding or replantingof hillslopes

The high degree of variabilityin the longevity and trajectory of the recovery curve is, in part, due tothe complexity of many interacting factors, including fire severity,catchment physiography, vegetation composition and regrowth, soils,geology, climate, site disturbance history, and post-fire land manage-ment (

Balfour,

The TRAPPI complex was unable to catalyze Ypt31/32 activation, suggesting that TRAPPII-specific subunits may contact Ypt31/32 to enable nucleotide exchange.

TRAPPII activates Ypt31/32 via the same active site residues used by TRAPPI to activate Ypt1, representing a remarkable case of active site plasticity

Three innovations distinguish our approach from previous investigations of TRAPPII activity: (1) we used synthetic liposome membranes to provide proper biological context; (2) we purified recombinant TRAPPII (rTRAPPII) from bacteria to avoid copurification of endogenous contaminants; and (3) we used enzymatic synthesis to generate native prenylated-Rab/GDP dissociation inhibitor (GDI) complex substrates.

In all of these studies, the observed TRAPPII GEF activity toward Ypt31/32 appeared low, likely because soluble Rab substrates were used, whereas in cells Rab GTPase activation occurs at the membrane surface of target organelles.

While the cause is unknown, itmight be the result of ectopic expression ofBdRAM1target genes and/or perhaps an interference of RAM1with other GRAS transcriptional networks, many ofwhich regulate development

hefinding thatB. distachyon ram1can support somefull arbuscule development suggests that other proteinsor pathways have the potential to compensate for lossofBdRAM1function.

it is noticeable thatBdRAM1transcriptlevels are low.

Arbuscules areephemeral structures, and the few arbuscules that areformed indellamutants display an increased lifespan,indicating that DELLA not only regulates arbusculeformation but also their degradation (

ram1

pproaches to increase the density of symbiotic interfaces will require a more focused, potentially cell type specificmanipulation of transcription factor gene expression.

However, the overexpressors also showaltered expression of hormone biosynthesis genes and aberrant growth patterns, including stunted bushy shoots and poor seedset

confers

As we have implied above, another intriguing puzzle to be solved is atwhat stage in evolution one group of theFlaviviridaedecided to restrict regulation oftranslation to a specific mechanism at the expense of another mechanism, and why?

a resultforwhich we have no explanation

Xrn-1

was strong under the control of the intergenic ZIKV or DENV 5=-UTRs

6Noncapped

Theresults showed that the nonmethylated (GpppA) RNAs generatedGlucexpression levelssurpassing the signal of the m7G-capped mRNA (Fig. 2B), a result suggesting that in thisexperiment,the classical m7G cap is not required for efficient translation

The qRT-PCRresultsshowed that the differences in translational efficiencies are not due to RNAinstability of the different reporter RNAs isolated from the transfected cells

DENV

D3

Gaussialuciferase (Gluc)

monocis-tronic

intragenic region (IGR

RESs)

we foundthat the y’’11fragment ion, which runs from Pro217 to Phe227and does not contain Thr216, is observed at a mass indicatingno phosphate. In contrast, the y’’12fragment ion, which runsfrom Thr216 to Phe227, is observed at a mass indicating thepresence of a phosphate adduct

Level of phosphorylation observed by semiquantitative MALDI mass spectrometry on synthetic peptides derived from HIV-1 RT after incubation with indicatedCDKs and ATP for 3 hr. +, phosphopeptide levels < 1%–25%; ++, phosphopeptide levels of 25%–50%; +++,phosphopeptide levels > 50% of total peptide signalintensity

a selective susceptibilityto phosphorylation by CDK2

hr216(corresponding to HIV-1 RT amino acid residues 210–227) wasobserved with both CDK2/cyclin A and CDK2/cyclin E, and tracephosphorylation was observed at two other peptides with CDK2/cyclin A (Figure 2A)

.

CDKs are requisite proline-directed serine/threonine kinases with preference for basicamino acids located C-terminal to the phosphate acceptor(Nigg, 1993). Therefore, the canonical CDK substrate motif isSer/Thr-Pro-X-Arg/Lys, where X can be any amino acid(s)

We hypothesizedthat CDK2 can support HIV-1 reverse transcription throughphosphorylation of HIV-1 RT

This was true for cellsinfected with R5-tropic (Figures 1G–1I) and with VSV-G-pseudo-typed (Figures 1J–1L) HIV-1.

CDK2, but not CDK7,resulted in a significant reduction of the proportion of HIV-1-pos-itive cells and of the frequency of per-cell levels of early, interme-diate, and late HIV-1 reverse transcripts.